The missing condition changed the decision

In 2021 a defined investment account had produced ₹6 crore of coupon receipts. Its unsold ₹100 crore portfolio had also risen in valuation to ₹104 crore. I combined the ₹6 crore receipt and ₹4 crore movement in a management paper and described ₹10 crore as cash supporting an ₹8 crore discretionary office-refit proposal. The arithmetic concealed the error because the total was easy to reproduce. No disposal had generated the additional ₹4 crore receipt implied by my description.

Treasury challenged the funding account before the recommendation was implemented. I withdrew it and supported a smaller ₹5.2 crore commitment considered against the identified ₹6 crore cash source. The ₹0.8 crore remainder described only that bounded source, not the insurer's overall liquidity or resources for claims. Permission to use funds, accounting treatment and capital and risk assessment remained separate professional questions. A cash receipt alone did not establish that an expense was authorised or that a distribution would be permissible.

We retained ₹21 lakh of separate expense for reviewing the presentation and its supporting process. That cost was not netted into the bounded refit calculation or described as money recovered. The revised proposal gave up part of the initial scope. Operating colleagues had to sequence the office work differently, and finance had to explain why a recommendation it had sponsored no longer stood on the stated funding basis. The case supplies no later investment sale or appreciation that conveniently repairs the original claim.

My correction was to make the state of an amount explicit before it enters a recommendation. Investment performance, recognised accounting movements and cash receipts can be related without being interchangeable. The financial paper must preserve the distinction appropriate to the decision. I keep the mistake in the profile because it shows how a senior finance leader can be misled by a correct total when the meaning of its components has been lost in the executive summary.

A 2022 software proposal quoted €1.8 million. The management plan translated that amount at ₹85 per euro, giving ₹15.3 crore. At an alternative planning rate of ₹95, the same foreign-currency amount would require ₹17.1 crore. The ₹1.8 crore difference came entirely from the translation assumption. A separate ₹1.1 crore rupee-denominated transition cost was also required. Neither rate is presented as a historical market observation or a prediction of the rate at which payment would occur.

The technology sponsor wanted to preserve the planned approval date because supplier and internal teams had reserved implementation time. I asked for currency and contractual review before the board treated ₹15.3 crore as a fixed commitment. We accepted a six-week delay to clarify the obligation, payment terms and treasury treatment. Technology colleagues retained the implementation assessment, while treasury and legal owners examined the financial and contractual questions. I did not substitute a preferred exchange-rate view for those responsibilities.

The management envelope approved for the proposal was ₹18.2 crore under the ₹95 planning scenario: ₹17.1 crore translated supplier amount plus ₹1.1 crore transition. That approval was not an executed hedge, a final payment or proof that no other contractual cost could arise. It described the scope and assumption on which the next decision proceeded. Any hedge or payment would require its own authorised process. The case contains no invented final rate or realised currency gain to make the outcome look successful.

The tradeoff was a slower approval and a less attractive rupee headline. I considered that preferable to asking implementation colleagues to absorb a currency difference that had never been made visible. The lesson was not to choose a permanently conservative rate for every proposal. It was to retain the currency of the obligation, explain the sensitivity and distinguish a planning envelope from the amount that will ultimately be settled. A quote can be commercially unchanged while the company's funding question changes substantially.

In 2023 a records-service proposal from an affiliate of the founder holding company quoted ₹5.6 crore annually. The defined account also required ₹1.4 crore of retrieval charges and ₹0.6 crore of internal oversight, bringing the annual comparison to ₹7.6 crore. An independent provider quoted ₹7.1 crore for the corresponding service scope, with ₹0.3 crore of retained oversight, giving ₹7.4 crore. The headline fee and the cost of the service arrangement were therefore different comparisons.

I recommended the unaffiliated option after the relevant operating, governance and professional reviews. The affiliate relationship required explicit assessment; a price comparison alone did not establish that terms were appropriate or that all related-party requirements had been met. Records owners and legal colleagues retained responsibility for access, protection and contractual arrangements. I held 0.8% within the employee ownership group, while the founder holding company owned 61% of the insurer. Those relationships remained disclosed rather than disappearing because finance preferred the unaffiliated quote.

The work incurred a separate ₹28 lakh one-time review expense. Including it, the selected first-year account was ₹7.68 crore for the stated scope. We did not turn the ₹0.2 crore annual model difference into a complete first-year saving or realised company-profit improvement. Future volume, retrieval needs and service performance could change the economics. The comparison also did not prove that the selected provider would deliver better records quality; that required its own operating evidence and supervision.

The difficult conversation concerned the status of the headline. A familiar supplier appeared less expensive until retained work was included. I wanted the board to see the same service perimeter on both sides before it considered price and relationship. The decision was a recommendation grounded in a fuller account, not an accusation that an affiliate's proposal was improper. I would repeat the separation of commercial arithmetic, service assurance and governance review, because allowing any one of those to stand in for the others makes the judgement less reliable.

Our 2024 management liquidity exercise used three illustrative additional-cash scenarios: ₹0 with a 40% assumed weight, ₹12 crore with 40% and ₹36 crore with 20%. The weighted mean was ₹12 crore. The initial paper placed that figure beside ₹18 crore of existing usable capacity and described a comfortable position. The mean did not establish that the company could meet the ₹36 crore scenario, which left an ₹18 crore gap in the same bounded exercise.

I asked the discussion to retain all three scenarios and the assumptions behind their weights. They were planning inputs, not proven probabilities, an actuarial claims estimate or a regulatory-capital calculation. Risk, actuarial, claims and investment professionals retained those separate assessments. The executive decision concerned how much liquidity to retain for the management exercise and what cost that choice would carry. I did not suggest that the largest of three invented cases represented the worst event the insurer could ever face.

The response was to retain an additional ₹18 crore of liquidity through changes to future investment maturities within the approved process. Using a 4% annual planning yield difference, the opportunity cost was ₹72 lakh a year. That was an assumption for comparison, not an observed income loss. We did not introduce an unapproved borrowing arrangement or claim that cash held for one purpose could automatically be used for another. Relevant permissions and the wider asset-and-obligation account remained part of the professional review.

The board saw the capacity decision and its cost together. The outcome did not demonstrate that every future demand could be met, or that the original mean was mathematically wrong. It showed that the mean answered a different question from capacity under the larger scenario. I wanted that distinction to survive the summary. A financial leader should be able to explain why retaining resources beyond an expected amount can be reasonable without implying that uncertainty has disappeared or that a particular scenario has been forecast with confidence.

A 2025 executive incentive proposal used a ₹3.6 crore maximum annual pool, with 80% weighting on premium growth and 20% on expense. The weights summed correctly, but they did not provide a sufficient account of the quality of the obligations being written or the service supporting them. I asked risk, actuarial and people colleagues to examine what the scorecard would encourage before a maximum budget became the basis for an implemented remuneration arrangement.

We developed a proposal with 40% weighting on premium growth, 30% on service quality and 30% on underwriting quality. The professional owners would need to define the quality measures and the evidence behind them. I did not choose a claims assumption or reserve estimate to produce a preferred incentive result. The proposal remained subject to board consideration and applicable policy, employment and regulatory review. It was not described as an adopted compliant scheme, a calculation of individual awards or a commitment to pay the full pool.

Design and simulation work cost ₹34 lakh, separately from the maximum incentive amount. That expense did not purchase proof that behaviour would improve. A poorly chosen quality measure can also create an incentive to optimise appearances, so the work had to consider how evidence would be reviewed and where professional judgement should remain distinct from reward. The company needed a defensible proposal rather than a more elaborate set of percentages that merely looked balanced on a slide.

The outcome recorded here is the revised proposal and the review still required. No payment or subsequent underwriting improvement is invented. The tradeoff was to make a simple volume-and-expense story more demanding before it could be used to reward executive performance. I wanted the maximum pool to remain a budget boundary, not an expected payout, and the measures to represent the institution's obligations more fully. Finance contributes by making those consequences visible while sharing design responsibility with the disciplines that understand quality and people decisions.